🍲 How I use AI to supercharge my portfolio career

Hello there 😎

In November 1958, on her 40th birthday, Madeleine L'Engle swore off writing for good.

She'd been at it for more than a decade, and the books still weren't paying the bills. She'd just opened another rejection letter. She covered the typewriter and cried as she paced around the room.

But writing still hadn't left her.

Even as she paced, she noticed her brain was still working out a story's plot. She couldn't turn it off.

So she uncovered the typewriter. She sat back down.

A few years later she published A Wrinkle in Time. More than two dozen publishers rejected it before it found a home, and then it won the Newbery Medal.

All told, she published more than sixty books: novels, poetry, essays, memoirs, Bible commentaries. In her spare time (hah!) she raised three kids, helped her husband run a general store in Connecticut, and spent nearly 30 years as the librarian at the Cathedral of St. John the Divine in New York.

How do you write sixty books around all of that? Here's what she said:

That was L'Engle's secret. Several pots.

I think it's the same secret to a portfolio career.

You keep a few things going. You drop something in one pot, walk past, drop something in another. Most days, nothing is ready. Then one day, you come back and you've got a fabulous stew waiting for you (or if you're Chinese, a mapo tofu).

That part hasn't changed. What's changed with AI is:

• How fast the pots cook

• How many pots one person can keep going at once

🔁 A portfolio career used to be one pot at a time

For a long time, I thought of a portfolio career as a five-step progression:

1/ Get good. Build a skill a company will actually pay for.

2/ Get effective. Do the same work in fewer hours, so you have room for anything else.

3/ Get a remote job. This gives you control over where you live and when you work.

4/ Start consulting. Use those extra hours to take on your first clients while you're still employed.

5/ Productize. Package what you learned from clients into something you can sell again and again. But only after you've sold the service, because the one-on-one work is where you become an expert.

Yes, you could skip steps. You just didn't want to, because each step builds on the one before it.

I said in a recent issue that this is no longer true. Here's the longer version of why:

Pre-AI, the thing that made a second pot expensive wasn't the work. It was the switch. Every time you moved from one project to another, the whole context had to come back up in your head: who it's for, what the rules are, where you left off.

That context reload is why "service first, productize later" was the right advice. Most people can't afford two context reloads a day, never mind five.

Now the context lives in markdown files. The goal, the audience, the rules you learned the hard way, all written down where the tools can read them. The reload is cheap.

Which means the real constraints on a portfolio career changed too. They used to be time and attention. Now they're 1/ how well you can use the tools and 2/ whether you actually execute.

🍳 How many pots can one person actually keep going?

Here's where I think most people (me included, until recently) still underestimate this.

When I read the L'Engle quote, I picture an old stove, like one you'd find in a Little House on the Prairie book. Two burners, max. That's one burner for your day job, and one for an intensive side project.

(I know what you're thinking: Ma cooked over an open fireplace in Little House on the Prairie, not a stove! This is true! But... later, in On the Banks of Plum Creek, Pa actually bought Ma an iron cookstove with a built-in oven. Pa was so thoughtful.)

That's pre-AI math. When you had to hold every pot's context in your own head, two was a lot. (OK, if you're really brain maxxxing, maybe four? But for the rest of us mere mortals, two was plenty.)

But if the tools hold the context for you, it's not a stove anymore. It's not two or three more burners. It's six pots. Eight. More than you'd have believed a single person could watch.

You're going from a two-burner cookstove to an industrial kitchen armed with six burners, a salamander, a deep fryer, and your mise en place station. Which looks something like this:

What my portfolio career looks like right now: two clients, three newsletters, a handful of small public web tools for Philly parents, the systems that run our house (dinner, money, taxes, payroll), and rewriting an old script. Two years ago that list would have been impossible.

Two things change when you take that seriously.

1/ Your bandwidth. You don't have to hold every detail of every project in your head. You have to know the shape of it and the next move. The details are in the pot.

2/ Your mindset. Years ago I wrote about splitting your day into a manager mindset and a maker mindset. The manager chunks the work down and sets the day. The maker does the work. I'd add now that the manager job got a lot bigger and the maker job got narrower. Most of your hours become manager hours: deciding what goes in which pot, checking what's done, sending the next thing. The maker hours get rarer, and it takes a deliberate effort to protect them.

⏳ This is a moment in time

I remember listening to Gary Vaynerchuk talk about Google Ads in its early days. He was building up his dad's liquor store and getting about 10 cents per lead. So he poured more and more of his budget there, and that's how he built their business.

Looking back, his regret was that even after adding more budget, he still didn't spend enough. He kept spreading his budget across other channels, too. He knew it was a gold rush, and he still hedged.

(Two caveats. There's obvious survivorship bias in a story like that. And with your personal finances, I firmly believe in diversifying. Time is different.)

I think about that story a lot with AI right now. If we fast-forward three years and look back, I think we'll see this as an unusual moment, for three reasons:

1/ The models are cheap. For what they can do, a working tool has never been this cheap. I doubt the monthly plans stay this generous for long.

2/ The gap is wide. The distance between someone who builds real things with these tools every day and someone who's tried a chatbot a few times has never been bigger. In three years, everyone will catch up.

3/ You have the time. If you have any control over your hours (a remote job, a client or two, a Saturday morning), you have something most people don't: room to play with the tools and make things.

If you accept those three, the question stops being "should I start a second pot?" It becomes:

1. What are all the things you could build?

2. Which ones first, and how many can run at once?

3. And what else should you be using AI for that you haven't yet, because it would unblock everything else?

🧭 How to manage multiple pots without losing your mind

Take all of this with a grain of salt. Everyone's different, and my version is probably overkill for 90% of people. Here are a few principles that hold up no matter how many pots you're running:

1/ Start the day with what's stuck, not what's due. Most of what blocks you isn't hard. It's an email you haven't answered or a decision you haven't made. I sat on a contractor's quote for a day and a half before I realized I could have it checked for fairness and request a second one in the time I spent avoiding it. Every morning, find the thing that hasn't moved and ask whether AI can move it one step.

2/ Ask "can AI help with this?" before "should I?" Renewing a car registration, planning the week's dinners, deciding between two job offers. The answer is yes more often than you'd think.

3/ Add pots slowly. Online poker players often "multi-table": two tables at first, then four, then eight. The big mistake is jumping up before you're ready. Start slow. Two projects at a time, try that for a week. When that gets boring, add a third.

4/ Know which mode you're in. A morning of fast context switching (i.e., manager mode) feels great. I feel superhuman. Then I try switching to maker mode in the afternoon and I simply can't do it; my brain is fried. You have to plan your modes, not just the specific tasks. To stretch L'Engle's metaphor further (her quote is doing a lot of work in this piece): some days are for moving pots. Some days are for standing over one. Decide which before the day starts.

5/ Give it time. The systems that carry my client work didn't appear. They built up, very slowly, over more than a year, from brain dumps to spreadsheets to saved workflows. If the second pot isn't paying by next month... all good. Value takes time to create.

🔥 You can do it all

A portfolio career used to happen one step at a time. Now you do it all. Maybe not all at once, but certainly in parallel.

It takes practice. So if you've been waiting to start a second thing until the first one was finished, fire up your favorite AI tool this weekend and let her rip.

🤝 When you’re ready, here’s how I help you land your next job:

I love meeting new people. Say hi on LinkedIn 👋